Multiagent Simulation for Generating Land-Use Policy and Planning Scenarios

Agent-based modeling is aimed at simulating individuals and their behavior. The outcome of agent actions has a direct impact on the surroundings and can then be analyzed in aggregate fashion. This project was aimed at a small local community of 3,000 residents in the city of Chelan, Chelan County, Washington. The goal was to simulate urban growth and land-use change in a 25-year period (2000–2025) based on calibrated historical data for 1990 and 2000 using the CommunityViz Policy Simulator extension.

The input data was used to generate a simulated community of individuals in Chelan. The agents, representing people and businesses of Chelan, lived and made individual decisions that had direct impact on their environment. The emphasis of the project was on the location of residential and business buildings under various policy scenarios.

The first scenario assumed that there were no major policy changes compared to the historical period. The zoning restrictions allowed for conditional switches from one land use to another. The second scenario imposed building permit restrictions beginning with 30 permits per year in scenario one and decreasing to 20 permits per year in scenario two. The third scenario addressed the issue of preserving the northern city outskirts from buildup. Therefore, selected northern parcels were excluded from development. Finally, scenario four implemented residential building density restrictions: no residential development was allowed outside the urban growth area, yet the density limit was increased from eight to 10 buildings per acre within this area. The biggest differences in building location and building density were observed between the first and the last scenarios.